杂波未知条件下基于箱粒子滤波的CBMeMBer算法

A CBMeMBer Algorithm Based on Box Particle in Unknown Clutter Environment

针对杂波未知条件下,传统的势均衡多目标多伯努利滤波器(CBMeMBer)的序贯蒙特卡洛实现跟踪精度不高,且所需粒子数目过大,导致跟踪效率低下的问题,引入区间分析理论,提出了杂波未知条件下基于箱粒子滤波技术的CBMeMBer算法。该算法构建目标和杂波的混合状态空间模型,基于箱粒子滤波技术,建立杂波模型,推导出目标预测、更新方程,用多目标箱粒子CBMeMBer递推表达式估计目标状态。仿真实验表明,在杂波模型先验已知或未知条件下,所提算法既保证了目标跟踪精度,又大幅度提高了算法的执行速率。

In unknown clutter environment,the traditional Sequential Monte Carlo(SMC)implementation of Cardinality Balanced multi-target multi-Bernoulli(CBMeMBer)filter cannot guarantee a good performance,and multitude number of particles leads to time consuming and low efficiency of tracking.Aiming at this problem,this paper introduced the theory of interval analysis,and proposed the CBMeMBer algorithm based on box particle filter in unknown clutter environment.Targets and clutter hybrid state space models were established,then,establishing clutter model and deriving the prediction equation and updating equation based on box particles.The state of multi-target was recursively estimated in utilization of CBMeMBer filter box particles.Simulation revealed that the proposed algorithm ensured tracking accuracy of target and greatly improved the algorithm's execution speed under the prior known or unknown conditions of the clutter model.